INS Projects (old)

See current research theme 'Multi-Sensor Integration'

The Integration of INS/DR with GPS, GLONASS and Pseudolites

Satellite-based kinematic positioning techniques such as those based on GPS or Glonass can offer consistent precision during the period of satellite signal tracking. However, such systems require line-of-sight between the orbiting satellite(s) and the receiver antenna(s). (The same holds true even when using pseudolites in place of, or as augmentations to, satellite signals.) In some situations, for example in highway tunnels or under trees, bridges, etc., receivers cannot track GPS/Glonass satellite (and/or pseudolite) ranging signals, and thus will fail to provide positioning results. In contrast, Inertial Navigation Systems (INS) are self-contained navigation sensors and do not require any external signals. Hence, it is well known that the integration of INS with GPS (and/or Glonass, and/or pseudolites) can overcome the disadvantages of the (pseudo-)satellite-based systems.

For some vehicle navigation applications, 3-axis INS systems are replaced by simpler single-axis gyro (usually based on Fibre Optic, or FOG, technology) and 'odometer' sensors. Collectively these systems are known as 'Dead Reckoning' systems (DR). One of the main drawbacks of an INS or DR when operated as a stand-alone system is the growth of systematic errors with time. Fortunately precise GPS, Glonass and/or pseudolite measurements are ideally suited for the 'calibration' (or estimation) of the INS/DR systematic errors. A calibrated INS or DR system can then provide high-rate precise positioning and attitude information. Therefore, the integration of INS or DR with technologies such as GPS, Glonass and/or pseudolites is an effective technique for many applications requiring precise positioning and orientation information.

The SNAP group at The University of New South Wales has as one of its long-standing objectives the development of a navigation 'platform' to support both fundamental and applied research in the area of integrated satellite/pseudolite and inertial navigation systems. Research efforts have been directed to (and will continue):

Pseudolite studies have progressed to the stage where software for combined GPS and pseudolite data processing (pseudo-range and carrier phase measurements) has been developed and tested on a variety of projects, using several pseudolite technologies (see Theme 4).

INS studies initially were focussed on understanding the measurements made by the solid-state MIGITS (Miniature Integrated GPS/INS Tactical System), jointly owned by UNSW and Curtin University of Technology. INS and GPS/INS integration research in the School is being directed by Dr. Jinling Wang. Assoc. Prof. Dorota Grejner-Brzezinska has made available Ohio State University's AIMS software to launch the GPS/INS integration project. Hung-Kyu Lee, commenced his PhS research on this topic in February 2001 and submitted his PhD for examination in early 2004.

Dead Reckoning (DR) investigations have essentially been concerned with collecting data from GPS, FOG and odometer sensors in the context of a 'train tracking' project at BHP's Port Kembla Steelworks. A system based on the above sensors, using a simple Kalman filter, was developed by Michael Moore, and its performance was compared with an 'off-the-shelf' Continuous Positioning System (CPS). Michael continued research into multi-antenna attitude determination systems.